0
Original Research: Cardiothoracic Surgery |

Predicting Mortality and Morbidity After Elective Cardiac Surgery Using Vasoactive and Inflammatory Biomarkers With and Without the EuroSCORE ModelVasoactive Biomarkers to Predict Mortality FREE TO VIEW

Abraham Schoe, MD; Emile F. Schippers, MD, PhD; Stefan Ebmeyer, MD; Joachim Struck, PhD; Robert J. M. Klautz, MD, PhD; Evert de Jonge, MD, PhD; Jaap T. van Dissel, MD, PhD
Author and Funding Information

From the Department of Intensive Care (Drs Schoe and de Jonge), Department of Infectious Diseases (Drs Schippers and van Dissel), and Department of Thoracic Surgery (Dr Klautz), Leiden University Medical Centre, Leiden, The Netherlands; and Thermo Fisher Scientific/BRAHMS GmbH (Drs Ebmeyer and Struck), Hennigsdorf, Germany.

CORRESPONDENCE TO: Abraham Schoe, MD, Department of Intensive Care, Leiden University Medical Centre, B4Q, Albunisdreef 2, PO Box 9600, 2300 RC Leiden, The Netherlands; e-mail: a.schoe@lumc.nl


FUNDING/SUPPORT: Biomarker values were determined courtesy of Thermo Fisher Scientific/BRAHMS GmbH.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2014;146(5):1310-1318. doi:10.1378/chest.13-2615
Text Size: A A A
Published online

BACKGROUND:  In cardiac surgery, preoperative mortality risk assessment tools like the EuroSCORE help to guide physicians in optimizing perioperative care of patients. We investigated the value of preoperative levels of inflammatory (procalcitonin [PCT]) and vasoactive (C-terminal pro-arginine vasopressin [CT-proAVP], midregional pro-atrial natriuretic peptide [MR-proANP], midregional proadrenomedullin [MR-proADM], and C-terminal pro-endothelin-1 [CT-proET-1]) biomarkers for risk assessment of mortality and morbidity and compared it with the EuroSCORE.

METHODS:  We performed a prospective observational cohort study in a single-center academic medical hospital and analyzed 746 consecutive patients undergoing elective cardiac surgery. In a directly preoperative blood sample, we assessed PCT, CT-proAVP, MR-proANP, MR-proADM, and CT-proET-1 levels.

RESULTS:  In single-variable logistic regression models, all biomarkers predicted 30-day mortality. The biomarkers CT-proET-1 (C statistic, 0.785; 95% CI, 0.687-0.883) and MR-proADM (C statistic, 0.780; 95% CI, 0.671-0.889) predicted 30-day mortality. For the EuroSCORE, the C statistic was 0.689 (95% CI, 0.594-0.784). There was a significant improvement in the prediction of 30-day mortality when the EuroSCORE was combined with MR-proADM (C statistic, 0.792; 95% CI, 0.699-0.884) or CT-proET-1 (C statistic, 0.798; 95% CI, 0.715-0.880). The model with EuroSCORE, MR-proADM, and CT-proET1 had the highest C statistic of 0.803 (95% CI, 0.717-0.890) and was significantly better than the EuroSCORE alone.

CONCLUSIONS:  In elective cardiac surgery, preoperative levels of MR-proADM and CT-proET-1 are predictors of 30-day mortality and could improve the predictive accuracy of the EuroSCORE. Further research should confirm the place of these new biomarkers in the prediction of mortality and identification of patients at risk.

Figures in this Article

Cardiac surgery risk assessment models like the EuroSCORE are widely used in predicting mortality after cardiac surgery and provide guidance in the clinical management of patients. Moreover, these models are used as a measure of quality of hospital care and even the performance of individual surgeons1,2 for funding allocation, patient counseling, and various scientific reasons.3,4 Given their broad application, risk assessments need to be continually optimized to keep them up to date. In this respect, there is a growing interest in the use of biomarkers as independent screening, diagnostic, and prognostic tools.5

In nonsurgical cardiovascular research, several studies have addressed the additive value of biomarkers to optimize risk prediction. For instance, Zethelius et al6 demonstrated that adding biomarkers to a risk model for cardiovascular death in elderly men improved the performance of the overall model as evidenced by a small, but significant increase of the C statistic (from 0.664 without biomarkers to 0.766 with all biomarkers included). However, another study using a different set of inflammatory biomarkers in a different patient population found a lack of improvement.7

Several assays for prohormones have been developed that might yield information about the pathophysiologic pathways involved in cardiovascular performance, such as the inflammatory pathway (procalcitonin [PCT]810), the vasomotor response (midregional proadrenomedullin [MR-proADM]11,12 and C-terminal pro-endothelin-1 [CT-proET-1]13,14), the neurohumoral response (C-terminal pro-arginine vasopressin [CT-proAVP]1517), and the cardiohumoral response (midregional pro-atrial natriuretic peptide [MR-proANP]1820). In the present study, we assayed the serum levels of these biomarkers prior to surgery and evaluated whether these molecules, singularly or combined, could predict mortality in patients undergoing elective cardiac surgery and whether they add to the predictive strength of the EuroSCORE.

Research Goals

In a prospective, observational cohort study in patients undergoing elective cardiac surgery, we investigated whether the biomarkers PCT, MR-proADM, CT-proET-1, CT-proAVP, and MR-proANP were associated with 30-day mortality and with the composite end point complicated outcome, which was defined as needing > 48 h of ICU care or as death in the hospital. Furthermore, in a logistic model, we evaluated whether the predictive value of the EuroSCORE could be improved with the addition of these biomarkers.

Study Population

The study was performed at the Leiden University Medical Centre, a 500-bed tertiary referral hospital. Study participants were consecutive patients undergoing elective cardiac surgery. Patients aged ≥ 18 years who could read and understand the informed consent and the study information were eligible for entrance in the study. We adhered to the Helsinki protocol and standard of good clinical practice. In all cases, written informed consent was obtained prior to enrollment, usually at the visit to the outpatient clinic the week before the operation. The study was approved by the Medical Ethical Committee (protocol P05.054). Exclusion criteria were pregnancy, active infection, emergency surgery, and participation in another study.

Operative Procedure

Anesthesia was standardized and consisted of premedication with lorazepam (1-1.5 mg) the night before and morning of the operation. Induction was achieved with propofol 1 to 1.5 mg/kg bolus and a remifentanyl 20 to 40 μg bolus. Anesthesia was maintained with propofol 5 to 6 mg/kg/h continuously and remifentanyl 800 to 1,600 μg/h continuously. Heparin was administered before the start of cardiopulmonary bypass, if applicable, as a 300 International Units/kg bolus and subsequently adapted to an activated clotting time > 400 s. At the end of cardiopulmonary bypass, heparin was antagonized with protamine sulfate titrated 1:1 to the initial dose of heparin (3 mg/kg). Ventilation was set at low pressure (bilevel pressure ventilation, 10/4 cm H2O) and low frequency to prevent atelectasis of the lung.

Bypass priming included 1,000 mL hydroxyethyl starch 130 (Voluven 6%; Fresenius Kabi Norge AS) and 200 mL Ringer’s solution, 100 mL mannitol 20%, 5,000 International Units heparin, 1,000 mg tranexamic acid, and 1,000 mg cefazolin. After venous and arterial cannulation, bypass was commenced using a heart-lung machine (S3; Sorin Group) with a centrifugal blood pump (Revolution; Sorin Group). Oxygenation was accomplished with a hollow fiber oxygenator (QUADROX-i; MAQUET Holding BV & Co KG). Tubing was coated with bioinert heparin-free polymers (SAFELINE; MAQUET Holding BV & Co KG). Flow was laminar, with rates set at 2.0 to 2.6 L/m2/min. Intermittent warm antegrade blood cardioplegia was instituted every 15 to 20 min for a period of 2 min at a flow of 300 to 450 mL/min with a pressure of 200 to 250 mm Hg. During bypass, core temperature was maintained at 34°C to 36°C. Standard surgical prophylaxis consisted of cefazolin 1,000 mg IV administered within 1 h of the first incision.

EuroSCORE

All patients received a EuroSCORE. Patients were scored according to the EuroSCORE 1 method described in Nashef et al.3

Biomarkers

A more detailed description of the assessed biomarkers (PCT, CT-proAVP, MR-proANP, MR-proADM, CT-ProET-1) can be found in e-Appendix 1. Analysis of biomarkers was done on coded samples in an external laboratory independent of the clinical team and without knowledge of clinical data at the courtesy of Thermo Fischer Scientific/BRAHMS GmbH. Briefly, MR-proADM, PCT, MR-proANP, and CT-proET-1 were all analyzed with the Thermo Fisher Scientific/BRAHMS KRYPTOR using time-resolved amplified cryptate emission technology, which is described in detail elsewhere.2123 CT-proAVP was analyzed with sandwich immunoluminometric assay developed by Thermo Fisher Scientific/BRAHMS GmbH.17 With these commercially available methods and equipment, clinicians and laboratories can readily sample and determine all biomarkers mentioned.

Baseline Examinations

After cannulation of the radial artery, but before anesthesia induction, an arterial blood sample (9 mL in Vacuette lithium-heparin collection tubes [Santa Cruz Biotechnology, Inc]) was drawn. The blood sample was immediately processed (centrifuged for 5 min at 4,000 rpm), and plasma was stored at −80°C until analysis. Patient characteristics were derived from electronic medical records. The EuroSCORE and operative data were derived from the database held by the department of cardiac surgery of the Leiden University Medical Centre. The EuroSCORE was calculated as defined by Nashef et al.3 Perioperative and postoperative data from the operating room and ICU were derived from the patient data management system used in the ICU (MetaVision; iMDsoft]. Mortality data were derived from the Dutch basic population registry if the patient did not die in the hospital.

Outcomes and Follow-up

The primary outcome was 30-day all-cause mortality. The secondary outcome measure was a complicated vs uncomplicated outcome after cardiac surgery. In this respect, a complicated outcome was defined as needing postoperative ICU care for > 48 h or death in the hospital from all causes. Patients were ready for discharge to the ward when they were on low oxygen (maximum, 5 L/min), low inotropes (maximum dobutamine, 4 μg/kg/min), and no vasopressors and were clinically stable.

Data Analysis and Statistics

Continuous variables were checked for normality. Variables with a nonnormal distribution (all biomarkers) were log10 normalized. Significance of differences of the mean between groups was tested using t test if the distribution of the data was normal. For dichotomized variables, we used Pearson χ2 test or Fisher exact test, where appropriate. Pearson χ2 test was used to compare categorical data. We did a complete case analysis so that all models would represent the same patient population and different models could be compared. This means that all patients who had a missing value for one or more biomarkers would be excluded from the analysis.

To analyze the additive value of individual biomarkers to the performance of the EuroSCORE, we used logistic regression and receiver operating characteristic (ROC) analysis with a C statistic. First, all biomarkers and the EuroSCORE were tested in a univariate model for prediction of 30-day mortality and a complicated outcome. ROC analysis was performed per biomarker and the area under the curve (C statistic) calculated. As a next step, various logistic models were made with the EuroSCORE and one or several biomarkers added to the model. With the predicted values of the models, we did an ROC analysis and calculated a C statistic (area under the curve). Where possible, we compared the C statistic of the various models with that of the EuroSCORE and calculated a P value. We considered the C statistics to be statistically significantly different if P ≤ .05.

Anticipating that few patients would meet the primary end point, we decided to do a sensitivity analysis for the most discriminating biomarker models. In this analysis, we constructed the same models, leaving out two deaths with the highest EuroSCORE, two deaths with the lowest EuroSCORE, two deaths with the highest biomarker level, and two deaths with the lowest biomarker level for the appropriate model. The same procedure was repeated with the difference of leaving out two survivors instead of deaths. We calculated the C statistics with 95% CIs for these models and compared them with the original model.

PCT was below the lower detection limit of 0.017 ng/mL in 13 of 717 cases. These cases were imputed with a value between 0 and the detection limit of 0.0085 ng/mL. Analyses were done with Stata/MP 12 (StataCorp LP) statistical software.

Patient Characteristics and Outcome

From December 2006 to August 2010, 817 patients gave written informed consent to participate in the study. In 17 patients, the preoperative sample was not drawn and, thus, could not be analyzed. One hundred twenty-one subjects had at least one missing preoperative biomarker value and were also excluded from analysis. The reasons for missing values were that the sample was not drawn at the correct time point or was lost during processing, transportation, or storage in the Leiden University Medical Centre or external laboratory. The collection process was at random and does not reflect problems with the actual biomarker analysis. A total of 679 patients (mean age, 65.7 ± 11.5 years; 458 men) were analyzed (Table 1).

Table Graphic Jump Location
TABLE 1 ]  Baseline Characteristics of All Patients

Data are presented as No. (%) or mean ± SD. The cumulative value of the surgical procedures exceeds the total patient number because of combinations between categories. Categorical variables are according to Nashef et al3 except for the LVF. EuroSCORE is calculated exactly according to Nashef et al.3 CABG = coronary artery bypass graft; EF = ejection fraction; LVF = left ventricular function; NYHA = New York Heart Association.

The primary end point of 30-day postoperative mortality was reached in 23 patients (3.4%). Overall, heart failure, compromised preoperative echocardiographic performance, pulmonary hypertension, and more complex surgery (combination valve/coronary artery bypass graft and surgery on myocardium) were significantly more present in the nonsurvivors than survivors. The mean ± SD EuroSCORE was higher among nonsurvivors than survivors (8.0 ± 3.3 vs 5.5 ± 2.9, respectively; P < .0005) (Table 2).

Table Graphic Jump Location
TABLE 2 ]  Baseline Characteristics per Outcome Group

Data are presented as No. (%) or mean ± SD unless otherwise indicated. Five patients could not be allocated to the complicated or uncomplicated group and were left out of the analysis. The cumulative value of the surgical procedures exceeds the total patient number because of combinations between categories. Categorical variables are according to Nashef et al3 except for the LVF. EuroSCORE is calculated exactly according to Nashef et al.3 See Table 1 legend for expansion of abbreviations.

a 

P < .05.

b 

P < .0005 by Pearson χ2 or Fisher exact test if expected value < 5.

c 

P < .05.

d 

P < .0005 by Pearson χ2 for the whole group.

e 

P < .0005 by t test.

The secondary composite end point of a complicated course was reached in 163 patients (27.5%). The median length of stay in the ICU was 0.98 day (range, 0.58-102.93 days), with a mean of 3.21 ± 8.30 days. Twenty patients were readmitted during their hospital stay or within 30 days, five of whom died. Diabetes, chronic kidney insufficiency, heart failure, worse preoperative echocardiographic findings, pulmonary hypertension, and more-complex surgery were significantly more present in patients with a complicated clinical course. The mean EuroSCORE was higher among patients with a complicated clinical course than in those with an uncomplicated clinical course (5.6 ± 2.9 vs 7.8 ± 3.0, respectively; P < .0005) (Table 2). During the ICU stay, which ranged between 3 h and 102 days in the whole population, 27 patients (3.6%) died, 21 of whom within 30 days postoperatively.

Biomarkers

Mean biomarker levels were significantly higher in 30-day nonsurvivors than in survivors and in the patients with a complicated clinical course than in those with an uncomplicated clinical course (Table 3). There was overlap between groups (Figs 1AE).

Table Graphic Jump Location
TABLE 3 ]  Biomarkers Geometric Mean and Median per Outcome Group

CT-proAVP = C-terminal pro-arginine vasopressin; CT-proET-1 = C-terminal pro-endothelin-1; MR-proADM = midregional proadrenomedullin; MR-proANP = midregional pro-atrial natriuretic peptide; PCT = procalcitonin.

a 

Significantly different within the outcome group, with P < .05 by t test on log10 normalized mean with equal variance for PCT, CT-proAVP, and MR-proANP and with unequal variance for CT-pro-ET-1 and MR-proADM. Five patients could not be allocated to the complicated or uncomplicated group and were left out of the analysis.

Figure Jump LinkFigure 1 –  Levels of prohormones by study outcome group. A, PCT levels. B, MR-proADM levels. C, CT-proET-1 levels. D, CT-proAVP levels. E, MR-proANP levels. The box represents the median and 25th and 75th percentiles. The whiskers represent the upper and lower limits of the adjacent value, meaning that every value within 1.5 times the interquartile range of the nearer quartile, as defined by Tukey.24 CT-proAVP = C-terminal pro-arginine vasopressin; CT-proET-1 = C-terminal pro-endothelin-1; MR-proADM = midregional proadrenomedullin; MR-proANP = midregional pro-atrial natriuretic peptide; PCT = procalcitonin.Grahic Jump Location

The C statistics of the EuroSCORE and biomarkers and combinations are summarized in Table 4. All biomarkers were associated with mortality and a complicated outcome in the univariate analysis. The C statistic of the EuroSCORE was 0.689 (95% CI, 0.594-0.784). MR-proADM and CT-proET-1 had the highest C statistic for 30-day mortality compared with other biomarkers and the EuroSCORE (0.780 [95% CI, 0.671-0.889] and 0.785 [95% CI, 0.687-0.883], respectively). PCT had the lowest C statistic (0.648; 95% CI, 0.534-0.762). Compared with the EuroSCORE, none of the biomarkers had a statistically significant higher C statistic. The results of the sensitivity analysis, as described in the Materials and Methods section are presented in e-Table 1, which shows that the C statistic varies but little compared with the original model. The lower boundaries of the 95% CI of all sensitivity models were > 0.500. Correlations between biomarker means are shown in Table 5. The highest correlation was found between MR-proADM and CT-proET-1 (r = 0.87).

Table Graphic Jump Location
TABLE 4 ]  Univariate and Adjusted Logistic Regression Models and Their AUCs (C Statistics) per Outcome Group

C statistics of the various models are compared with the C statistic of the EuroSCORE. AUC = area under the curve. See Table 3 legend for expansion of abbreviations.

a 

For difference in C statistic compared with EuroSCORE.

b 

With EuroSCORE and indicated biomarker entered in the model.

Table Graphic Jump Location
TABLE 5 ]  Correlation Coefficients Among Biomarkers

See Table 3 legend for expansion of abbreviations.

Biomarkers Added to the EuroSCORE

When the EuroSCORE and separate biomarkers were modeled together, the C statistic of the separate models was significantly higher than that of the EuroSCORE alone for 30-day mortality (except for PCT modeled with the EuroSCORE). The models with CT-proET-1 and MR-proADM had the highest C statistic (Table 4). The C statistics did not improve much more with different combinations of biomarkers with or without the EuroSCORE entered into the model (data shown only for the best model with the EuroSCORE MR-proADM and CT-proET-1). The prediction models for the secondary end point, a complicated clinical course, had a low C statistic compared with the models with 30-day mortality as the end point. Although some combined models with a biomarker and the EuroSCORE were significantly higher than the EuroSCORE alone, their C statistics were low (Table 4).

The main finding of the present study is that a single preoperative determination of the prohormones MR-proADM and CT-proET-1, and to a lesser extent CT-proAVP and MR-proANP, could add significantly to the overall predictive performance of the commonly applied EuroSCORE. In contrast, both preoperatively determined biomarkers and the EuroSCORE were less able to predict a complicated course defined by an ICU stay > 48 h or overall all-cause in-hospital mortality.

There are several potential weaknesses and strengths of this study. First, in this cohort, the EuroSCORE had a rather disappointing C statistic of 0.689, which was less than expected. In the original article establishing the EuroSCORE, a C statistic of 0.759 in the validation cohort was reached. One explanation could be that the present single-center academic study population is different with respect to the cohort used by Nashef et al.3 At the Leiden University Medical Centre, heart failure surgery represents a relatively large proportion of the patients, and this kind of surgery may be overrepresented in the present study. Second, because of the limited number of events in the 30-day mortality group, we had to be prudent with respect to the number of parameters that could be entered into the logistic models because of the danger of overfitting. Finally, the secondary end point of complicated course has its limitations. We chose this end point because it was readily available and, in our opinion, in line with the pathophysiologically plausible premise that if a biomarker represents a decompensated state, the patient would recover more slowly after surgical trauma or not survive the hospital stay. If a biomarker could indeed predict slower recovery or death, it would be of great value to clinicians and decision-makers. A drawback is that the current end point is not precise and heterogeneous, that is, some complications occurred > 48 h later or necessitated readmission, points that we did not focus on. A composite end point of major adverse cardiac events or other specified complications might have been better and should be investigated in future studies alongside the current end points.

One of the strengths of this study is the relatively large size of the population, and with 697 subjects, the findings are robust, as the sensitivity analysis shows. Leaving out patients from whom a large impact on the C statistic of the predictive model would be expected showed, in fact, only a small change and did not affect the conclusions. Second, the study focused on a well-recognizable group of patients undergoing elective cardiac surgery; thus, excluding emergency surgery and reducing patient heterogeneity allowed for the generalization of the findings and will enable a comparison with future studies from other heart centers. Finally, in this observational study, we did not interfere with perioperative procedures or care, and the biomarker results were not available to the attending physicians. Thus, the data reflect the real-life care found in many hospitals.

From a pathophysiologic perspective, it is plausible that biomarkers, which give information about important pathways in cardiovascular disease, could be of value in predicting outcome in cardiac surgery. Adrenomedullin causes vasodilation and hypotension, increases cardiac output, and causes natriuresis,25 whereas endothelin-1 is a potent vasoconstrictor, which causes elevation of BP and a decrease of cardiac output by lowering stroke volume and frequency in healthy volunteers.13,26 Endothelin-1 and adrenomedullin seem to play an important, counteractive role in basal vasomotor homeostasis.27 There was a high correlation between MR-proADM and CT-proET-1 (r = 0.82) in the present study, which may underline the intricately regulated, counteractive nature of the two vasoactive hormones.

The EuroSCORE is elaborate, taking time and effort to gather all the data required. Moreover, the gathered data are as objective as possible but not immune to error and interpretation. It would be a step forward in risk assessment and mortality prediction if a reliable, affordable, and reproducible rapid blood test could achieve equal or better discrimination. More study in this field is needed.

It is important to note that we made a comparison with the EuroSCORE and not with the EuroSCORE II, which was launched in 2011 and published in 2012,28 because the present study stopped with patient recruitment in August 2010, and we did not have the acquired data to calculate the EuroSCORE II in retrospect. How the new and old EuroSCORE behave compared with each other and in various patient populations is still a matter of debate.2931 The main idea of the present study that a biomarker could be of aid in predicting mortality, therefore, remains valid. A future study should take the newer version of the EuroSCORE into account.

Biomarkers have become an adjunct to classic risk determinants and could help in screening, diagnosis, therapy, and stratification. The use of biomarkers and their role in risk assessment in patients undergoing elective cardiac surgery should be studied in larger cohorts and with a different patient mix to confirm the present, promising findings.

Author contributions: A. S. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. A. S. contributed to the study design and writing of the manuscript; E. F. S. contributed to the study design, data acquisition, and review and revision of the manuscript; S. E. and J. S. contributed to the biomarker data through BRAHMS GmbH (they had no influence on the study design, no access to clinical data or data analysis), and reviewed the manuscript regarding only the biomarker data; R. J. M. K. contributed to the data acquisition and writing and revision of the manuscript; E. d. J. contributed to the data analysis, interpretation of all statistical analyses, data acquisition, and writing and revision of the manuscript; and J. T. v. D. contributed to the study design, data analysis, interpretation of all statistical analyses, and writing and revision of the manuscript.

Financial/nonfinancial disclosures: The authors have reported to CHEST the following conflicts of interest: Drs Ebmeyer and Struck were both affiliated with Thermo Fisher Scientific/BRAHMS GmbH at the time of the study. They had, however, nothing to do with the design of the study or the collection and analysis of the data. Dr Klautz has received speaker fees and has been on the advisory board for Medtronic. Drs Schoe, Schippers, de Jonge, and van Dissel have reported that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Role of sponsors: Thermo Fisher Scientific/BRAHMS GmbH did not have an influence on the study design and had no access to clinical data or was involved in the data analysis or interpretation.

Other contributions: The authors thank R. Wolterbeek, MSc (Department of Medical Statistics, Leiden University) and E. van Es, Ing (perfusionist, Department of Thoracic Surgery, Leiden University Medical Centre).

Additional information: The e-Appendix and e-Table can be found in the Supplemental Materials section of the online article.

CT-proAVP

C-terminal pro-arginine vasopressin

CT-proET-1

C-terminal pro-endothelin-1

MR-proADM

midregional proadrenomedullin

MR-proANP

midregional pro-atrial natriuretic peptide

PCT

procalcitonin

ROC

receiver operating characteristic

New York State Department of Health. Coronary Artery Bypass Surgery in New York State 1996-1998. Albany, NY: New York State Department of Health; 2001.
 
New York State Department of Health. Adult Cardiac Surgery in New York State 2003-2005. Albany, NY: New York State Department of Health; 2008.
 
Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg. 1999;16(1):9-13. [CrossRef] [PubMed]
 
Nilsson J, Algotsson L, Höglund P, Lührs C, Brandt J. Comparison of 19 pre-operative risk stratification models in open-heart surgery. Eur Heart J. 2006;27(7):867-874. [CrossRef] [PubMed]
 
Braunwald E. Biomarkers in heart failure. N Engl J Med. 2008;358(20):2148-2159. [CrossRef] [PubMed]
 
Zethelius B, Berglund L, Sundström J, et al. Use of multiple biomarkers to improve the prediction of death from cardiovascular causes. N Engl J Med. 2008;358(20):2107-2116. [CrossRef] [PubMed]
 
Wang TJ, Gona P, Larson MG, et al. Multiple biomarkers for the prediction of first major cardiovascular events and death. N Engl J Med. 2006;355(25):2631-2639. [CrossRef] [PubMed]
 
Becker KL, Nylén ES, White JC, Müller B, Snider RHJ Jr. Clinical review 167: procalcitonin and the calcitonin gene family of peptides in inflammation, infection, and sepsis: a journey from calcitonin back to its precursors. J Clin Endocrinol Metab. 2004;89(4):1512-1525. [CrossRef] [PubMed]
 
Aouifi A, Piriou V, Bastien O, et al. Usefulness of procalcitonin for diagnosis of infection in cardiac surgical patients. Crit Care Med. 2000;28(9):3171-3176. [CrossRef] [PubMed]
 
Madershahian N, Wittwer T, Strauch J, et al. Kinetic of procalcitonin in the early postoperative course following heart transplantation. J Card Surg. 2008;23(5):468-473. [CrossRef] [PubMed]
 
Ishimitsu T, Ono H, Minami J, Matsuoka H. Pathophysiologic and therapeutic implications of adrenomedullin in cardiovascular disorders. Pharmacol Ther. 2006;111(3):909-927. [CrossRef] [PubMed]
 
Morgenthaler NG, Struck J, Alonso C, Bergmann A. Measurement of midregional proadrenomedullin in plasma with an immunoluminometric assay. Clin Chem. 2005;51(10):1823-1829. [CrossRef] [PubMed]
 
Böhm F, Pernow J. The importance of endothelin-1 for vascular dysfunction in cardiovascular disease. Cardiovasc Res. 2007;76(1):8-18. [CrossRef] [PubMed]
 
Papassotiriou J, Morgenthaler NG, Struck J, Alonso C, Bergmann A. Immunoluminometric assay for measurement of the C-terminal endothelin-1 precursor fragment in human plasma. Clin Chem. 2006;52(6):1144-1151. [CrossRef] [PubMed]
 
Jochberger S, Velik-Salchner C, Mayr VD, et al. The vasopressin and copeptin response in patients with vasodilatory shock after cardiac surgery: a prospective, controlled study. Intensive Care Med. 2009;35(3):489-497. [CrossRef] [PubMed]
 
Goldsmith SR. The role of vasopressin in congestive heart failure. Cleve Clin J Med. 2006;73(suppl 3):S19-S23. [CrossRef] [PubMed]
 
Morgenthaler NG, Struck J, Alonso C, Bergmann A. Assay for the measurement of copeptin, a stable peptide derived from the precursor of vasopressin. Clin Chem. 2006;52(1):112-119. [CrossRef] [PubMed]
 
McGrath MF, de Bold AJ. Determinants of natriuretic peptide gene expression. Peptides. 2005;26(6):933-943. [CrossRef] [PubMed]
 
Khan SQ, Dhillon O, Kelly D, et al. Plasma N-terminal B-type natriuretic peptide as an indicator of long-term survival after acute myocardial infarction: comparison with plasma midregional pro-atrial natriuretic peptide: the LAMP (Leicester Acute Myocardial Infarction Peptide) study. J Am Coll Cardiol. 2008;51(19):1857-1864. [CrossRef] [PubMed]
 
Morgenthaler NG, Struck J, Thomas B, Bergmann A. Immunoluminometric assay for the midregion of pro-atrial natriuretic peptide in human plasma. Clin Chem. 2004;50(1):234-236. [CrossRef] [PubMed]
 
Mathis G. Rare earth cryptates and homogeneous fluoroimmunoassays with human sera. Clin Chem. 1993;39(9):1953-1959. [PubMed]
 
Steinbach G, Rau B, Debard AL, et al. Multicenter evaluation of a new immunoassay for procalcitonin measurement on the Kryptor System. Clin Chem Lab Med. 2004;42(4):440-449. [CrossRef] [PubMed]
 
Caruhel P, Mazier C, Kunde J, Morgenthaler NG, Darbouret B. Homogeneous time-resolved fluoroimmunoassay for the measurement of midregional proadrenomedullin in plasma on the fully automated system B.R.A.H.M.S KRYPTOR. Clin Biochem. 2009;42(7-8):725-728. [CrossRef] [PubMed]
 
Tukey JW. Exploratory Data Analysis. Reading, MA: Addison-Wesley; 1977.
 
Kitamura K, Kangawa K, Eto T. Adrenomedullin and PAMP: discovery, structures, and cardiovascular functions. Microsc Res Tech. 2002;57(1):3-13. [CrossRef] [PubMed]
 
Weitzberg E, Ahlborg G, Lundberg JM. Differences in vascular effects and removal of endothelin-1 in human lung, brain, and skeletal muscle. Clin Physiol. 1993;13(6):653-662. [CrossRef] [PubMed]
 
Krzemiński K, Cybulski G, Ziemba A, Nazar K. Cardiovascular and hormonal responses to static handgrip in young and older healthy men. Eur J Appl Physiol. 2012;112(4):1315-1325. [CrossRef] [PubMed]
 
Nashef SA, Roques F, Sharples LD, et al. EuroSCORE II. Eur J Cardiothorac Surg. 2012;41(4):734-744. [CrossRef] [PubMed]
 
Biancari F, Vasques F, Mikkola R, Martin M, Lahtinen J, Heikkinen J. Validation of EuroSCORE II in patients undergoing coronary artery bypass surgery. Ann Thorac Surg. 2012;93(6):1930-1935. [CrossRef] [PubMed]
 
Carnero-Alcázar M, Silva Guisasola JA, Reguillo Lacruz FJ, et al. Validation of EuroSCORE II on a single-centre 3800 patient cohort. Interact Cardiovasc Thorac Surg. 2013;16(3):293-300. [CrossRef] [PubMed]
 
Barili F, Pacini D, Capo A, et al. Does EuroSCORE II perform better than its original versions? A multicentre validation study. Eur Heart J. 2013;34(1):22-29. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1 –  Levels of prohormones by study outcome group. A, PCT levels. B, MR-proADM levels. C, CT-proET-1 levels. D, CT-proAVP levels. E, MR-proANP levels. The box represents the median and 25th and 75th percentiles. The whiskers represent the upper and lower limits of the adjacent value, meaning that every value within 1.5 times the interquartile range of the nearer quartile, as defined by Tukey.24 CT-proAVP = C-terminal pro-arginine vasopressin; CT-proET-1 = C-terminal pro-endothelin-1; MR-proADM = midregional proadrenomedullin; MR-proANP = midregional pro-atrial natriuretic peptide; PCT = procalcitonin.Grahic Jump Location

Tables

Table Graphic Jump Location
TABLE 1 ]  Baseline Characteristics of All Patients

Data are presented as No. (%) or mean ± SD. The cumulative value of the surgical procedures exceeds the total patient number because of combinations between categories. Categorical variables are according to Nashef et al3 except for the LVF. EuroSCORE is calculated exactly according to Nashef et al.3 CABG = coronary artery bypass graft; EF = ejection fraction; LVF = left ventricular function; NYHA = New York Heart Association.

Table Graphic Jump Location
TABLE 2 ]  Baseline Characteristics per Outcome Group

Data are presented as No. (%) or mean ± SD unless otherwise indicated. Five patients could not be allocated to the complicated or uncomplicated group and were left out of the analysis. The cumulative value of the surgical procedures exceeds the total patient number because of combinations between categories. Categorical variables are according to Nashef et al3 except for the LVF. EuroSCORE is calculated exactly according to Nashef et al.3 See Table 1 legend for expansion of abbreviations.

a 

P < .05.

b 

P < .0005 by Pearson χ2 or Fisher exact test if expected value < 5.

c 

P < .05.

d 

P < .0005 by Pearson χ2 for the whole group.

e 

P < .0005 by t test.

Table Graphic Jump Location
TABLE 3 ]  Biomarkers Geometric Mean and Median per Outcome Group

CT-proAVP = C-terminal pro-arginine vasopressin; CT-proET-1 = C-terminal pro-endothelin-1; MR-proADM = midregional proadrenomedullin; MR-proANP = midregional pro-atrial natriuretic peptide; PCT = procalcitonin.

a 

Significantly different within the outcome group, with P < .05 by t test on log10 normalized mean with equal variance for PCT, CT-proAVP, and MR-proANP and with unequal variance for CT-pro-ET-1 and MR-proADM. Five patients could not be allocated to the complicated or uncomplicated group and were left out of the analysis.

Table Graphic Jump Location
TABLE 4 ]  Univariate and Adjusted Logistic Regression Models and Their AUCs (C Statistics) per Outcome Group

C statistics of the various models are compared with the C statistic of the EuroSCORE. AUC = area under the curve. See Table 3 legend for expansion of abbreviations.

a 

For difference in C statistic compared with EuroSCORE.

b 

With EuroSCORE and indicated biomarker entered in the model.

Table Graphic Jump Location
TABLE 5 ]  Correlation Coefficients Among Biomarkers

See Table 3 legend for expansion of abbreviations.

References

New York State Department of Health. Coronary Artery Bypass Surgery in New York State 1996-1998. Albany, NY: New York State Department of Health; 2001.
 
New York State Department of Health. Adult Cardiac Surgery in New York State 2003-2005. Albany, NY: New York State Department of Health; 2008.
 
Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg. 1999;16(1):9-13. [CrossRef] [PubMed]
 
Nilsson J, Algotsson L, Höglund P, Lührs C, Brandt J. Comparison of 19 pre-operative risk stratification models in open-heart surgery. Eur Heart J. 2006;27(7):867-874. [CrossRef] [PubMed]
 
Braunwald E. Biomarkers in heart failure. N Engl J Med. 2008;358(20):2148-2159. [CrossRef] [PubMed]
 
Zethelius B, Berglund L, Sundström J, et al. Use of multiple biomarkers to improve the prediction of death from cardiovascular causes. N Engl J Med. 2008;358(20):2107-2116. [CrossRef] [PubMed]
 
Wang TJ, Gona P, Larson MG, et al. Multiple biomarkers for the prediction of first major cardiovascular events and death. N Engl J Med. 2006;355(25):2631-2639. [CrossRef] [PubMed]
 
Becker KL, Nylén ES, White JC, Müller B, Snider RHJ Jr. Clinical review 167: procalcitonin and the calcitonin gene family of peptides in inflammation, infection, and sepsis: a journey from calcitonin back to its precursors. J Clin Endocrinol Metab. 2004;89(4):1512-1525. [CrossRef] [PubMed]
 
Aouifi A, Piriou V, Bastien O, et al. Usefulness of procalcitonin for diagnosis of infection in cardiac surgical patients. Crit Care Med. 2000;28(9):3171-3176. [CrossRef] [PubMed]
 
Madershahian N, Wittwer T, Strauch J, et al. Kinetic of procalcitonin in the early postoperative course following heart transplantation. J Card Surg. 2008;23(5):468-473. [CrossRef] [PubMed]
 
Ishimitsu T, Ono H, Minami J, Matsuoka H. Pathophysiologic and therapeutic implications of adrenomedullin in cardiovascular disorders. Pharmacol Ther. 2006;111(3):909-927. [CrossRef] [PubMed]
 
Morgenthaler NG, Struck J, Alonso C, Bergmann A. Measurement of midregional proadrenomedullin in plasma with an immunoluminometric assay. Clin Chem. 2005;51(10):1823-1829. [CrossRef] [PubMed]
 
Böhm F, Pernow J. The importance of endothelin-1 for vascular dysfunction in cardiovascular disease. Cardiovasc Res. 2007;76(1):8-18. [CrossRef] [PubMed]
 
Papassotiriou J, Morgenthaler NG, Struck J, Alonso C, Bergmann A. Immunoluminometric assay for measurement of the C-terminal endothelin-1 precursor fragment in human plasma. Clin Chem. 2006;52(6):1144-1151. [CrossRef] [PubMed]
 
Jochberger S, Velik-Salchner C, Mayr VD, et al. The vasopressin and copeptin response in patients with vasodilatory shock after cardiac surgery: a prospective, controlled study. Intensive Care Med. 2009;35(3):489-497. [CrossRef] [PubMed]
 
Goldsmith SR. The role of vasopressin in congestive heart failure. Cleve Clin J Med. 2006;73(suppl 3):S19-S23. [CrossRef] [PubMed]
 
Morgenthaler NG, Struck J, Alonso C, Bergmann A. Assay for the measurement of copeptin, a stable peptide derived from the precursor of vasopressin. Clin Chem. 2006;52(1):112-119. [CrossRef] [PubMed]
 
McGrath MF, de Bold AJ. Determinants of natriuretic peptide gene expression. Peptides. 2005;26(6):933-943. [CrossRef] [PubMed]
 
Khan SQ, Dhillon O, Kelly D, et al. Plasma N-terminal B-type natriuretic peptide as an indicator of long-term survival after acute myocardial infarction: comparison with plasma midregional pro-atrial natriuretic peptide: the LAMP (Leicester Acute Myocardial Infarction Peptide) study. J Am Coll Cardiol. 2008;51(19):1857-1864. [CrossRef] [PubMed]
 
Morgenthaler NG, Struck J, Thomas B, Bergmann A. Immunoluminometric assay for the midregion of pro-atrial natriuretic peptide in human plasma. Clin Chem. 2004;50(1):234-236. [CrossRef] [PubMed]
 
Mathis G. Rare earth cryptates and homogeneous fluoroimmunoassays with human sera. Clin Chem. 1993;39(9):1953-1959. [PubMed]
 
Steinbach G, Rau B, Debard AL, et al. Multicenter evaluation of a new immunoassay for procalcitonin measurement on the Kryptor System. Clin Chem Lab Med. 2004;42(4):440-449. [CrossRef] [PubMed]
 
Caruhel P, Mazier C, Kunde J, Morgenthaler NG, Darbouret B. Homogeneous time-resolved fluoroimmunoassay for the measurement of midregional proadrenomedullin in plasma on the fully automated system B.R.A.H.M.S KRYPTOR. Clin Biochem. 2009;42(7-8):725-728. [CrossRef] [PubMed]
 
Tukey JW. Exploratory Data Analysis. Reading, MA: Addison-Wesley; 1977.
 
Kitamura K, Kangawa K, Eto T. Adrenomedullin and PAMP: discovery, structures, and cardiovascular functions. Microsc Res Tech. 2002;57(1):3-13. [CrossRef] [PubMed]
 
Weitzberg E, Ahlborg G, Lundberg JM. Differences in vascular effects and removal of endothelin-1 in human lung, brain, and skeletal muscle. Clin Physiol. 1993;13(6):653-662. [CrossRef] [PubMed]
 
Krzemiński K, Cybulski G, Ziemba A, Nazar K. Cardiovascular and hormonal responses to static handgrip in young and older healthy men. Eur J Appl Physiol. 2012;112(4):1315-1325. [CrossRef] [PubMed]
 
Nashef SA, Roques F, Sharples LD, et al. EuroSCORE II. Eur J Cardiothorac Surg. 2012;41(4):734-744. [CrossRef] [PubMed]
 
Biancari F, Vasques F, Mikkola R, Martin M, Lahtinen J, Heikkinen J. Validation of EuroSCORE II in patients undergoing coronary artery bypass surgery. Ann Thorac Surg. 2012;93(6):1930-1935. [CrossRef] [PubMed]
 
Carnero-Alcázar M, Silva Guisasola JA, Reguillo Lacruz FJ, et al. Validation of EuroSCORE II on a single-centre 3800 patient cohort. Interact Cardiovasc Thorac Surg. 2013;16(3):293-300. [CrossRef] [PubMed]
 
Barili F, Pacini D, Capo A, et al. Does EuroSCORE II perform better than its original versions? A multicentre validation study. Eur Heart J. 2013;34(1):22-29. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Supporting Data

Online Supplement

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
Guidelines
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543